Urban road traffic congestion prediction based on knowledge graph

Urban road traffic congestion prediction holds significant research importance as it enables timely response to traffic flow patterns, optimizes urban planning, and enhances transportation efficiency. In order to improve the accuracy and recall of predicting urban road traffic congestion, a knowledge graph-based intelligent prediction approach is provided in this paper. This method first analyzes the measurement indicators of urban road traffic congestion status, and uses vehicle speed as the evaluation standard to develop a measurement standard for congestion degree. Secondly, a spatiotemporal knowledge graph of urban transportation networks was constructed using multi-source spatiotemporal data. Finally, based on the constructed spatiotemporal knowledge graph and the correlation of dynamic traffic, a novel prediction model for urban road traffic congestion was constructed using graph convolutional neural networks to obtain traffic congestion prediction results. The experimental results validate the effectiveness of the proposed approach, and indicate its significant potential for practical applications.